Antibodies are powerful and versatile binding molecules that the immune system employs to eliminate foreign antigens and as such constitute an excellent model for elucidating the principles governing macromolecular recognition. A knowledge of the molecular basis of antibody function is also essential for an in-depth understanding of immune responses and for the development of antibodies as therapeutic tools (e.g. through the "humanization" of rodent antibodies). We propose to explore the molecular basis of antigen-antibody recognition through detailed structure-function studies of site-directed mutants of the anti-hen egg-white lysozyme (HEL) monoclonal antibody D1.3, which has been the subject of extensive structural studies in our laboratory. For these experiments, the protein-engineered, bacterially-expressed FvDl.3 fragment will be used since the three-dimensional structures of the free Fv and that of the Fv-HEL complex are known to 1.8 Angstrom resolution; this should allow a more rigorous interpretation of site-directed mutagenesis experiments than is currently possible for any other antigen- antibody system. Fv mutants designed to investigate particular aspects of the problem of protein-protein (antigen-antibody) recognition will be characterized in terms of affinity and reactivity towards the antigen. Selected FvD1.3 mutants complexed with HEL will be crystallized and X-ray crystallographic analysis will be carried out to precisely ascertain the effects of particular amino acid substitutions at the structural level. In addition, titration calorimetry will be used to determine the entropy and enthalpy changes of the binding reactions with the aim of correlating these thermodynamic parameters with the X-ray models. To investigate how affinity maturation towards a defined protein antigenic determinant is achieved by the immune system, random mutagenesis of antibody genes in combination with recently developed methods for the display of antibodies on the surface of filamentous bacteriophage will be used to mimic this process and to isolate variants of D1.3 with improved affinity for antigen. The three-dimensional structures of the corresponding Fv fragments complexed with HEL will then be determined in order to understand the structural basis for the observed increases in affinity. It is envisaged that this basic work will generate a library of Fv mutants of known structure and affinity which will serve as a basis for modelling the structural and thermodynamic parameters of antigen-antibody reactions. This should significantly contribute to efforts aimed at ab initio prediction of antibody affinity and combining site conformation, and thereby facilitate the engineering of antibody molecules for medical and chemical applications.